Did low interest rates boost households’ consumption?

Marco Di Maggio, Amir Kermani, Rodney Ramcharan

05 October 2014

a

A

A major question facing the Federal Reserve, as well as other central banks, is whether the unconventional monetary policy measures adopted after the Crisis have succeeded in boosting aggregate demand. Six years after the Crisis, we still observe many debt-burdened households, widespread unemployment, and relatively low growth -- despite the highly expansionary monetary policy adopted by the Federal Reserve. Therefore, it is not a surprise that there is disagreement among economists, as well as policymakers, about the effectiveness of the monetary policy measures implemented after the financial crisis. While many economists believe that a decline in interest rates can have large effects on households’ consumption through either a decrease in household borrowing costs or an increase in their wealth, many others are sceptical of this, noting that credit market frictions and already high debt levels might limit the ability of households to take advantage of low interest rates.

A reason for this disagreement is that empirical tests aimed at measuring the impact of interest rate movements on households’ consumption remain elusive. The empirical challenge is that a household’s decision to refinance its mortgage might be correlated with unobserved household or local geographic characteristics that also determine spending behaviour. For instance, households with a bad credit history may be unable to refinance their mortgage, as banks may be unwilling to refinance them; the same may apply to liquidity-constrained households, which cannot pay the closing costs of their pre-existing mortgage.fn1 Similarly, households living in counties where the housing market has experienced a more severe crash are less likely to refinance their mortgages due to the high loan-to-value ratios, as most banks require 20% equity in the house. Hence, these households' consumption decisions might be driven more by local factors than any drop in interest rates.

The existing literature has focused mainly on the effects of a decline in interest rates on firms and banks. For instance, existing studies have investigated the channels through which monetary policy impacts banks' lending decisions and risk-taking behaviour (Jimenez et al. 2014, Jimenez et al. 2012 and Maddaloni and Peydró 2011). Recent studies on the households’ consumption behaviour during the Crisis include, among others, Mian and Sufi (2014), which examines the households' borrowing and spending behaviour resulting from rising house prices from 2002 to 2006; Mian and Sufi (2013), which investigates the elasticity of consumption with respect to housing net worth during the 2006-2009 period, and Parker et al. (2013), which investigates the effect of the Economic Stimulus Act of 2008 on households’ consumption. However, the impact of monetary policy on households' consumption behaviour at the disaggregated level remains mostly understudied.

Using hybrid adjustable rate mortgages to identify the effect of decline in interest rates

To overcome the aforementioned identification difficulties, our recent work has exploited anticipated changes in monthly payments of borrowers with adjustable rate mortgages originated between 2005 and 2007 (Di Maggio et al. 2014). These adjustable mortgages have an interest rate fixed for the first five years, and then automatically adjust at the end of this initial period. These types of mortgage contracts provide us with a source of quasi-random variation in the households’ ability to take advantage of the lower interest rates. In fact, a drop in interest rates benefits borrowers holding these hybrid adjustable mortgages without any need to refinance, simply due to the provisions of their mortgage contract. Moreover, this is a significant reduction because the adjustable mortgages originated in 2005 benefited from an average reduction of 3.3 percentage points in the reference interest rate in 2010.2

Among the sample of contracts that we study, the timing of the interest rate adjustment varies across households, and key for our identification strategy is to exploit the variation in the timing of the interest rate adjustment to study how households respond to these changes in the debt service. Also, by comparing households holding a relatively homogenous set of contracts apart from the timing of the adjustment, we can further reduce the potential for any unobserved borrower heterogeneity that might independently affect the refinancing decision, or the choice of the mortgage type and consumption behaviour.

Households’ response to the decline in interest rates

First of all, we show that the change in the mortgage monthly payments after the interest rate adjustment has been sizeable. Figure 1 plots the mortgage monthly payments for the four quarters before the adjustment, which we normalised to zero, and the two years after the interest rate adjustment. It shows that at the moment of the interest rate adjustment the monthly payment due by the average borrower falls by $900, resulting in an increase in income totalling in the order of tens of thousands of dollars over the remaining life of the mortgage.

Figure 1. OLS estimates of monthly mortgage payment with quarterly dummies around the interest rate adjustment  

Note: All regressions include households, time, and origination cohort-time fixed effects. Standard errors are heteroskedasticity robust and clustered at the household level. The 95% confidence intervals are drawn around the point estimates.

Next, we analyse the households’ consumption and saving behaviour after the interest rate adjustment. In the wake of the reduction in the monthly payment, the average household increases monthly car purchases by 40% (or equivalently $150 per month). Figure 2 plots the monthly expenditures toward car purchases in response to the interest rate adjustment. Since the change in the mortgage payments was anticipated, we observe a slight but statistically significant increase in the quarter before the change, but the households' consumption expenditures spike in the quarter after the reset and remain significantly higher in the subsequent two years. Moreover, the probability of purchasing a car increases by at least 45% after the interest rate reset.3

Figure 2. OLS estimates of monthly car purchases with quarterly dummies around the interest rate adjustment  

Note: All regressions include households, time, and origination cohort-time fixed effects. Standard errors are heteroskedasticity robust and clustered at the household level. The 95% confidence intervals are drawn around the point estimates.

Interestingly, we find no evidence of intertemporal substitution. In the quarters before the interest rate adjustment, households did not decrease their consumption, nor do we find that the effect is short-lived. Instead, it increases over time. Similar results are obtained with our second measure of consumption derived from credit cards.

These results seem to suggest that the decline in interest rates significantly benefited the economy by boosting consumption, confirming the view that aggressively reducing interest rates might be an important palliative after the collapse of a credit boom in real estate. Of course, credit booms can leave households highly leveraged once asset prices collapse and the desire to deleverage might attenuate somewhat the expansionary effects of monetary policy. Figure 3 plots the monthly mortgage prepayment in response to the interest rate adjustment. Households more than double their efforts to deleverage compared to the pre-adjustment period by employing on average 15% of the positive income shock to repay their mortgage faster.

Figure 3. OLS estimates of monthly mortgage prepayment with quarterly dummies around the interest rate adjustment  

Note: All regressions include households, time, and origination cohort-time fixed effects. Standard errors are heteroskedasticity robust and clustered at the household level. The 95% confidence intervals are drawn around the point estimates.

We complement the previous findings by showing that the marginal propensity to consume is significantly higher for liquidity-constrained borrowers, as measured by their income and for those that had experienced a larger decline in housing wealth, as captured by their current loan-to-value ratio.

Debt rigidity and aggregate effects

If borrowers' marginal propensity to consume is higher than that of lenders, a decline in interest rates can result in a positive income shock that should increase consumption and boost economic activity. To estimate the effect of monetary policy on the county-level consumption, we exploit the geographical variation in the presence of adjustable-rate mortgages. As shown by Figure 4, adjustable mortgages have been more popular in some regions in the US rather than others. Specifically, zip codes in coastal areas have experienced higher levels of adjustable mortgages origination than other regions.

Figure 4. The fraction of adjustable-rate mortgages originated in each county in 2006

Then, by looking at quarterly car sales between 2007 and 2013, we show that changes in the interest rate tend to have a disproportionately larger effect on durable goods consumption in counties with a greater fraction of adjustable rate mortgages. Moreover, these same counties experience significantly higher deleveraging rates than counties with larger share of fixed-rate mortgages. This evidence suggests that debt rigidity – namely the responsiveness of loan contracts to interest rate changes -- plays an important role in the transmission of monetary policy to the real economy.

Conclusions

Bernanke's announcement in June of last year that the Fed would scale back on its quantitative easing has been greeted with anxiety by the markets and the media. Our results can inform the discussion on the adverse effects that the exit from quantitative easing could have on aggregate consumption. Moreover, these findings also point out that interest rate or monthly payment reductions can help improve households’ cash-flow and liquidity constraint problems, and should be considered as potential policy responses in the aftermath of asset price booms fuelled by debt.4

References

Agarwal, S, J C Driscoll, and D I Laibson (2013), “Optimal mortgage refinancing: A closed-form solution”, Journal of Money, Credit and Banking 45(4), 591.622.

Di Maggio, M, A Kermani and R Ramcharan (2014), “Monetary Policy Pass-Through: Household Consumption and Voluntary Deleveraging”, Columbia Business School Research Paper No. 14-24. Available at SSRN: http://ssrn.com/abstract=2489793.

Eberly, J and K Arvind (2014), “Efficient Credit Policies in a Housing Debt Crisis”, Stanford University working paper.

Fuster, A and P S Willen (2013), “Payment Size, Negative Equity, and Mortgage Default”, National Bureau of Economic Research working paper.

Hurst, E and F Stafford (2004), “Home is Where the Equity is: Mortgage Refinancing and Household Consumption”, Journal of Money, Credit and Banking, 985-1014.

Jimenez, G and S Ongena (2012), “Credit Supply and Monetary Policy: Identifying the Bank Balance-Sheet Channel with Loan Applications”, The American Economic Review 102(5), 2301-2326.

Jimenez, G, S Ongena, J L Peydro, and J Saurina (2014), “Hazardous Times for Monetary Policy: What do Twenty-Three Million Bank Loans Say about the Effects of Monetary Policy on Credit Risk-Taking?” Econometrica 82(2), 463-505.

Maddaloni, A and J L Peydró (2011), “Bank Risk-Taking, Securitization, Supervision, and Low Interest Rates: Evidence from the Euro-Area and the US Lending Standards”, Review of Financial Studies 24(6), 2121-2165.

Mian, A, K Rao, and A Sufi. (2013), “Household Balance Sheets, Consumption, and the Economic Slump”, The Quarterly Journal of Economics 128(4), 1687-1726.

Mian, A R and A Sufi. (2014), “House Price Gains and U.S. Household Spending from 2002 to 2006”, Fama-Miller Working Paper, Available at SSRN: http://ssrn.com/abstract=2412263.

Parker, J A, N S Souleles, D S Johnson, and R McClelland (2013), “Consumer Spending and the Economic Stimulus Payments of 2008”, The American Economic Review 103(6), 2530-2553.

Footnotes

1 See Hurst and Stafford (2004) and Agarwal et al. (2013) on these issues.

2 Fuster and Willen (2013) have shown that the borrowers holding these hybrid adjustable rate mortgages exhibit significantly lower default rates than other borrowers with similar characteristics after the adjustment.

3 We can compare our estimates with the existing literature on households’ consumption response to income shocks. Among the most recent contributions to this literature, Parker et al. (2013) analyse the households’ consumption reaction to the Economic Stimulus Act (ESA) of 2008. They find that households spent about 12 to 30% of their stimulus payments on nondurable consumption goods, and another 38 to 60% on the purchase of vehicles. On the one hand, we find a smaller effect if you compare the $150 spent on vehicle with the estimates provided by Parker et al. (2013). On the other, we find that due to the very different source of the income shock these effects last for up to two years after the interest rate adjustment, which makes the overall consumption spending significantly larger. Another element to be considered when making this comparison is that interest paid on a mortgage is tax deductible, which makes the effective income shock about 30% lower, depending on the tax brackets.

4 See Eberly and Krishnamurthy (2014) for a theoretical examination of this point.

a

A

Topics:  Global crisis Monetary policy

Tags:  adjustable rate mortgages, low interest rate, household consumption, deleveraging

Assistant Professor of Business Administration, Harvard Business School

Assistant Professor at the Haas School of Business and Department of Economics, UC Berkeley

Associate Professor of Public Policy, Price School of Public Policy; Associate Professor of Finance and Business Economics, Marshall School of Business, University of Southern California